US Surge Protector trips RCD

[UKMeterman]

Ah, if it is a double pole switch and it closes live first then the whole lot will be at 230V above earth and the surge protector will then provide a path live to earth. Also the transformer will take a while to settle down and could provide some interesting spikes.

 

[Megawatt]

If you want a better long-term solution you could get a non-protected US distribution strip and add a UK style surge protector to the autotransformer input.

With the 18th edition of the IET wiring regulations surge protection devices (SPD) at the supply to a home became a factor to consider but I suspect very few home will have that for years to come.

All new homes built or service changes will have to have SPD’s located at the main panel starting this year according to the 2020 edition of the NEC

 

[pc1966]

All new homes built or service changes will have to have SPD’s located at the main panel starting this year according to the 2020 edition of the NEC

That is good to hear, basically same ethos as our regulations!

While most mains stuff should be OK with 4kV spikes a lot of consumer electronics is crap and so it not

 

[Lucien Nunes]

I like the DP switch explanation, and given your observation that it only trips at switch-on, either that or the inrush transient is almost certainly the cause.

RCD's dont like capacitors to ground.

In the absence of any reference to the size of capacitors, I think that's too broad a generalisation. Many appliances and electronic goods have capacitors to ground and are intended to operate from an RCD. Looking at my bench equipment, I probably have 40 separate devices with Y-caps to ground all running from the one RCD, IIRC the leakage is around 8-10mA. A standard power strip made for the USA market would be intended to run from a 6mA GFCI which may trip from 4mA upwards, so would only be permitted to leak a fraction of that current.

However, you may well have a point that in this case the presence of the capacitors is increasing the leakage pulse due to the transient.